Spatially Resolved Optical Emission Spectroscopy in Plasma Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current plasma optical emission spectroscopy techniques in semiconductor processing primarily capture data from a single elongated volume, leading to averaging of plasma optical emission spectra and loss of local variations in chemical species concentrations, making it difficult to determine two-dimensional distributions necessary for uniform processing outcomes.
Innovation Solution
A method and system that measure plasma optical emission spectra along multiple non-coincident rays using a controller to select an optical intensity distribution function comprising basis functions that vary with both radial and circumferential locations, allowing for the determination of spatial distributions of plasma optical emission and chemical species concentrations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If optical emission spectra are collected from a single elongated volume, then the measurement system is simple and easy to operate, but local variations of plasma optical emission spectra and chemical species concentrations are lost due to averaging
Solution Approach 1:
The optical emission measurement is segmented into multiple discrete rays rather than collecting from a single elongated volume. The optical system divides the plasma measurement space into N distinct line-of-sight paths, each providing independent spectral data from a specific spatial location, thereby preserving local variations while maintaining system simplicity
Solution Approach 2:
The measurement approach transitions from one-dimensional spatial averaging along a single ray to two-dimensional spatial mapping through multiple rays. By arranging N rays in different positions and orientations across the plasma cross-section, the system captures emission variations in both radial and azimuthal directions, adding spatial dimensionality to the measurement
2Loss of information
If multiple non-coincident rays are used to measure plasma optical emission, then two-dimensional distributions of chemical species can be determined, but the optical measurement system becomes more complex
Solution Approach 1:
A single multi-channel spectrometer performs multiple measurement functions simultaneously by receiving optical signals from N different rays. The universal optical system can measure emission spectra from various spatial locations without requiring separate spectrometers for each ray, reducing overall system complexity while capturing comprehensive spatial information
Solution Approach 2:
Optical fibers serve as intermediaries to transmit optical emission signals from multiple ray locations to the spectrometer. These flexible waveguides enable the collection of spatially distributed emission data without requiring direct optical access from each measurement point to the spectrometer, simplifying the optical system architecture
3Measurement precision
If tomographic inversion techniques are used to determine spatial distribution from integrated measurements, then spatial distribution can be reconstructed, but the technique requires a large amount of acquired data and is computationally intensive
Solution Approach 1:
The patent extracts spatial distribution information directly from the multi-ray optical emission spectra without applying complex tomographic inversion algorithms. By measuring emission at multiple discrete rays and using curve fitting with basis functions, the method extracts spatial variations directly from the spectral data, eliminating the need for computationally intensive reconstruction techniques
4Reliability
If a single elongated volume is used for optical emission measurement, then the optical system is simple, but endpoint detection across the entire substrate surface is limited
Solution Approach 1:
The measurement system transitions from single-point spectral measurement to two-dimensional spatial mapping through multiple rays. This dimensional expansion enables endpoint detection across the entire substrate surface by capturing emission variations in both radial and azimuthal directions, providing comprehensive monitoring for reliable endpoint determination
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables the capture of two-dimensional distributions of plasma optical emission spectra and chemical species concentrations across the substrate, improving process uniformity and endpoint detection in plasma processing.
Implementation Method 1
Optical emission spectroscopy (OES) has proven itself as a useful tool for process development and monitoring in plasma processing. In optical emission spectroscopy, the presence and concentrations of certain chemical species of particular interest, such as radicals, is deduced from acquired optical (i.e. light) emission spectra of the plasma
Data Source
AI summary
Disclosed is a method, computer method, system, and apparatus for measuring two-dimensional distributions of optical emissions from a plasma in a semiconductor plasma processing chamber. The acquired two-dimensional distributions of plasma optical emissions can be used to infer the two-dimensional distributions of concentrations of certain chemical species of interest that are present in the plasma, and thus provide a useful tool for process development and also for new and improved processing tool development. The disclosed technique is computationally simple and inexpensive, and involves the use of an expansion of the assumed optical intensity distribution into a sum of basis functions that allow for circumferential variation of optical intensity. An example of suitable basis functions are Zernike polynomials.